Astrocytes– more than just passive scaffolding cells

Historically, astrocytes were often considered as passive scaffolding cells of the mammalian CNS. However, work over the past 2 decades suggests that astrocytes play a more active role and are involved in a wide variety of complex and essential functions in the healthy brain. Astrocytes are specialized glial cells that are distributed throughout the CNS.

According to their cellular morphology and location, astrocytes can be divided into two main classes, the protoplasmic and fibrous astrocytes. The latter are located in white matter tracts and are characterized by long, sparsely branched processes contacting nodes of Ranvier.

Protoplasmic astrocytes, which are found in grey matter regions, have many branching processes, which envelop synapses (Cajal, 1909). As heterogenous as they are there are even more astrocytic subpopulations including Bergmann glia in the cerebellum, Müller glia in the retina, pituicytes in the neurohypophysis, cribrosocytes at the optic nerve head, and others.

Astrocytic endfeet project directly to blood vessels, therewith participating in the formation and maintenance of the blood-brain barrier and providing access to nutrients from the cerebral blood stream (Kacem et al., 1998), whose flow velocity can be adapted by astrocytes in response to changes in neuronal activity (Attwell et al., 2010; Gordon et al., 2007). Moreover, astrocytes are interconnected among each other in a coordinated syncytium and are coupled to oligodendrocytes via gap junctions (Giaume et al., 1991; Nagy et al., 2003). In this manner a highly organized network between neurons, astrocytes, oligodendrocytes and capillaries is generated that allows bidirectional communication and may provide routes for metabolites.

The relationship between astrocytes and neurons, assured by the intimate physical connection of astrocytic processes to synapses, starts already during neurogenesis when astrocytes guide neuronal migration, survival and process extension. Later they are involved in formation, maintenance and remodelling of synapses mainly through release of trophic factors such as brain-derived neurotrophic factor (Powell et al., 1999; Ullian et al., 2001a).

There is growing evidence that in adulthood astrocytes are active partners of synapses. They express a plethory of transporters important for the clearance of neurotransmitters from the synaptic cleft (Genoud et al., 2006). Moreover, they were reported to be responsible for the recycling of synaptically released glutamate and GABA through the glutamate/GABA-glutamine cycle (Bak et al., 2006; Rothstein et al., 1996). A variety of receptors are expressed on astrocytes, which response to neuronal activity with an increase of intracellular calcium concentrations (Dani et al., 1992; Nimmerjahn et al., 2009). A specific consequence of astrocytic internal calcium elevations is the secretion of so-called gliotransmitters, including glutamate, adenosine triphosphate, GABA and d-serine (Parpura et al., 1994;

Zhuang et al., 2010; Panatier et al., 2006), which adapts synaptic transmission and plasticity (Gourine et al., 2010; Chen et al., 2012). This often results in several, often opposite types of effects, including stimulation or inhibition of synaptic transmission and participation in long-term potentiation or depression (Panatier et al., 2011; Yang et al., 2003; Shigetomi et al., 2013). The underlying mechanism how astrocytes release transmitters still remains a subject of debate (e.g. reviewed by Hamilton and Attwell, 2010; Araque et al., 2014). However, these findings have led to the concept of “tripartite synapse”, which represents a functional view of synaptic physiology that considers astrocytes as active contributors controlling neuronal information transfer.

Concomitantly, another model has flourished implying that astrocytes also support brain activity by supplying neurons with energy metabolites. This concept was originally proposed by Magistretti and Pellerin in 1994 claiming “Glutamate uptake into astrocytes stimulates aerobic glycolysis” (Pellerin and Magistretti, 1994a). This publication could be considered as the cornerstone for the beginning of a new research field and the development of the

“Astrocyte-to-Neuron-Lactate-Shuttle” (ANLS) hypothesis. Here, they described a tight metabolic connection between cortical synapses and their surrounding glial cell, in which the activity-dependent glutamate release in the synaptic cleft at glutamatergic synapses in the cortex is followed by the clearance of glutamate via glutamate transporters on astrocytes.

The entry of sodium that iscotransported with glutamate activates astrocytic Na+/K+ATPase, whose activity results in an enhancement of the glycolytic flux, hence the glucose uptake from the capillaries is stimulated. Lactate, the major end product of aerobic glycolysis, is released by astrocytes and taken up by neurons where it can be further metabolized in the tricarboxylic acid (TCA) cycle, thereby contributing to the energy budget of neurons (Pellerin and Magistretti, 1994). The rapid exchange of metabolites, such as lactate across the plasma membrane of cells requires expression of appropriate transporters. Monocarboxylate transporters (MCT) catalyze the proton-linked transport of glycolysis products pyruvate or lactate (Pierre and Pellerin, 2005). Three isoforms (MCT1, 2, and 4) are expressed in the CNS, whose distribution is heterogenous (Halestrap and Wilson, 2012). MCT1 and MCT4 were shown to be expressed by astrocytes, whereas MCT2 is preferentially expressed by neurons (Pellerin et al., 1998; Pierre et al., 2000, 2002), which is one prerequisite for lactate shuttling from astrocytes to neurons. Importantly, the disruption of actrocyte- or neuron-specific MCTs affects long-term memory in vivo, suggesting a trophically, supportive role of astrocytes for neuronal function (Suzuki et al., 2011).

Fig. 4 Schematic representation of the ANLS hypothesis

Glutamate (Glu) release at the active synapse stimulates neuronal glutamatergic receptors (GluR). A large proportion of the glutamate is taken up by astrocytes via excitatory amino acid transporters (EAATs, more specifically GLT-1 and GLAST) together with sodium ions (Na⁺). Na⁺ is extruded by the Na⁺/K⁺ ATPase, consuming ATP which triggers nonoxidative glucose utilization in astrocytes and glucose uptake from blood vessels through the glucose transporter GLUT1. Pyruvate is converted to lactate by the lactate dehydrogenase 5 (LDH5) and shuttled to neurons via monocarboxylate transporters (MCT1 and MCT4 in astrocytes, MCT2 in neurons). In neurons, this lactate is metabolized to pyruvate (Pyr) by LDH1 and used as a metabolite to support the neuronal energy budget.

Concomitantly, astrocytes participate in the recycling of synaptic glutamate by its conversion to glutamine (gln) by the glutamine synthetase (GS) and its subsequent transport to neurons, where it is converted back to glutamate by glutaminase (GLS). This figure is taken from Bélanger et al., 2011.

The concept of the ANLS based on several studies revealing striking metabolic differences between astrocytes and neurons. That neurons consume most of the energy during brain activation was already discovered in 1977 by positron emission tomography (PET) imaging of labelled F-fluoro-2-deoxyglucose (Sokoloff et al., 1977). Several PET studies in awake adult humans by Fox and Raichle, led to a fundamental rethinking of brain metabolism. They detected that the activity-dependent increases in blood flow and glucose uptake were only partly matched by parallel raises in oxygen utilization (Fox et al., 1988; Fox and Raichle, 1986). These investigations strongly indicate that neuronal activity stimulates aerobic glycolysis. Nonetheless, due to resolution limitations the cellular contribution remained elusive. However, studies on transcriptomic level of individually isolated cells revealed a different metabolic profile of neuronal and astrocytic cells indicating a prevalence of glycolytic pathways in astrocytes (Lovatt et al., 2007; Cahoy et al., 2008). In accordance with this, investigations of Itoh and Bouzier-Sore confirmed a higher glycolytic capacity in astrocytes

when compared to neurons in which oxidative metabolism predominates. Further, they figured out that astroglia metabolize glucose mainly to lactate which is released into the extracellular space and preferentially taken up and oxidized by neurons over pyruvate/

lactate produced intracellularly by glycolysis (Itoh et al., 2003; Bouzier-Sore et al., 2006).

This lactate utilization was directly assessed in human brains by MRS imaging of ¹³C-labeled lactate and was detected to be neuron specific (Boumezbeur et al., 2010). More important, lactate is able to maintain neuronal activity in vivo and even a preference of neurons to lactate over glucose in the presence of both metabolites is observed (Wyss et al., 2011).

However, the cellular origin and its possible contribution to the energy metabolism of the CNS are still issues of controversial debate. By now, direct in vivo evidence for the ANLS hypothesis is elusive and emphasizes the importance to further investigate the metabolic interactions between astrocytes and neurons.

2.2 T

W

– PKM2